Discussion Forum

Hello:

I have been sending an email to comsol support address and it returns a message saying that the server is down or doesn´t have enough space for files ...I want to contact Mr. Vallejo or somebody from Somsol whom can take a look at my cases , where I made some modifications explained in my failed email as follow. (though I couldn´t attach more than one file in this forum)

Dear Mr. Vallejos:

I modified my simulation by using an ambient temperature (Tamb) instead of heat flux. The boundary 21 in the model is set to convective and radiation heat flux , qo=0 . I try 2 ways to run the solver.


1) transient analysis inputting a Temperature polynomial, the case solution looks very similar to my parametric study. However , I would like your help double checking that my boundary condition is properly set!!! I think is good but have no performed a validation, where I could compare with previous work. I have been looking for analysis and papers , but I have not found some where the bc varies with time.


2) a parametric solver, in order to facilitate the nonlinear effects (i use the nonlinear solvers in both cases). In this case, I was able to set my temperature from a minimum of 298.15 to 316.7 K which is the max. Since the span is smaller , it should be easier than using q0 values (which peaks at 800 w-m2)...so the case works but it does no represents a day temperature profile . at minimum I would like to be able to set a cosine shape with the same max and min values, I try to apply such function in the parametric settings but it does not work. Is there a way to make tvar (my parameter) a cosine shape , at least???? I read that using parametric analysis may replace the transient case by using very small increments in the chosen parameter. in my case it would help me to compare even if the temperature profile is not exactly the same, or is there a way I could make it equal?

I understood that when the boundary has very sharp changes , COMSOL make no be able to converge, so I´m trying to find a way to smoothly get to my solution. for a 24 hours period. Please see my excel file where only temperature or heat flux can actually represent my bounday over time.

I will create a different case using a slighly different way to input Cp(T)

I would appreciate your advise on each case. I am attaching both runs. transient and parametric , and mi bc file.

Thanks

Virginia-